6. PCR-DGGE Characterization of Bacterial Associations from Soil Microcosms Arrays
6.2 Results and Discussion
6.2.1 Analysis of bacterial community profiles from soil microcosm arrays at two
6.2.1.2 Bacterial community diversity
0 5 10 15 20 25 30 35 40 45 50
0 20 40 60 80 100
Sampling Time (Weeks)
M e a n S p e c ie s R ic h n e s s
Figure 6.5 Changes in the Bacterial species richness (S) over depth and time. Symbols correspond to destructively sampled soil arrays A (◊), B (■), C (∆), and D (●) at times (weeks): 12 (T1); 32 (T2); 52 (T3); and 80 (T4), respectively. (a) Treatment HLRh and (b) Treatment HLRl.
Figure 6.6 Changes in the mean species richness measured at weeks 12 (T1), 32 (T2), 52 (T3), and 80 (T4) in soil arrays A, B, C, and D, respectively. Symbols correspond to treatments HLRh (▲) and HLRl (■).
0 10 20 30 40 50 60
0 5 10 15 20 25 30 35 40
0 10 20 30 40 50 60
0 5 10 15 20 25 30 35 40
Species Richness (S)
Depth (cm) (a)
(b)
The mean S recorded for treatment HLRh at T1 was significantly different from that recorded at T2 and T3 (F=<0.001) but not significantly different from that recorded at T4 (Table 6.2). In contrast, mean S recorded for treatment HLRl was significantly different (F=<0.001) at all sampling times (Table 6.2). Mean S was highest in array B, followed by arrays C, D, and A for both treatments.
Bacterial growth and incidence of organic contamination share a close correlation (Arora, Linde, Revil, and Castermant, 2007). The periodic supply of leachate to each set of arrays ensured a supply of substrates required for bacterial growth. This then appeared to lead to an increase in the overall S for both treatments (T1-T3). However it is feasible that eventually, the supply of phenol containing leachate became toxic leading to a decrease in S as observed in treatment HLRh. The non significant change in S observed at T1 and T4 supports this explanation. The difference in S observed for treatment HLRl at T1 and T4 remained significant by virtue of a larger difference in S, lending further support to the theory of toxicity under higher leachate loading rates.
Table 6.2 General Analysis of Variance comparing the mean Species Richness (S), Shannon- Weaver Index (H’), and the Shannon-Weaver Evenness Index (E) over time, and for two different hydraulic loading rates (HLR).
Mean Measures of Diversity Treatment Sampling
Times S H E
Time 1 18.19a 2.750a 0.9502a
Time 2 38.06c 3.407c 0.9402ab Time 3 27.94b 3.105b 0.9417ab
HLRh
Time 4 19.25a 2.683a 0.9157b
<0.001 <0.001 0.067
3.750 0.1669 0.02649
F-value l.s.d
s.e.d 1.875 0.0835 0.01325
Time 1 19.56a 2.779a 0.9409a
Time 2 44.50d 3.540d 0.9341a
Time 3 35.50c 3.344c 0.9410a
HLRl
Time 4 24.81b 2.968b 0.9306a
<0.001 <0.001 0.856
3.930 0.1688 0.02887
F-value l.s.d
s.e.d 1.964 0.0844 0.01444
The mean S recorded along the vertical profile of the soil arrays were compared for both leachate treatments at the four sampling intervals (Figure 6.6). Treatment HLRl showed consistently higher mean S over the four sampling times. Both treatments followed the same trend regarding S, an initial increase in mean S, peaking at T2, followed by a steady decrease at T4. The difference in mean S recorded for both treatments at T1 was not significant (P=0.283), however significant differences between the treatments were found at T2 (P=0.002), T3 (P=0.002), and T4 (P=<0.001) by virtue of the pairwise-t-test.
This adds to the body of evidence supporting the conclusion that pollutants supplied at a higher HLR become increasingly toxic to the existent Bacterial communities, thereby affecting the number of surviving/thriving species (S). A band (species) initially detected using PCR-DGGE may not be adequately amplified to generate an intensity significant enough to warrant detection in a sample later on in the investigation, even though that band may still be present (Jackson et al., 2001). In this investigation the number of samples made it impractical for comparison of replicated samples over time since samples were analyzed on different DGGE gels over the course of the investigation.
6.2.1.2.2 Shannon-Weaver index of diversity (H’)
The influence of soil depth on H’ for both leachate loading rates is shown in Figure 6.7. All soil arrays of both treatments follow similar trends with depth, with the major difference being the change in H’ at each depth over the four sampling times. The highest H’ values at each depth were produced by array B (week 32) of both treatments, subsequently followed by arrays C (week 52), A (week 12), and D (week 80) for treatment HLRh and arrays C, D, and A for treatment HLRl.
Redox conditions within each soil array were found to be dynamic, constantly changing over time. This was attributed to change in chemical and microbial compliments.
This constant state of flux could therefore be expected to affect the metabolic response of the resident microorganisms. Indeed, Løngborg et.al. (2006) concluded that the rate of xenobiotic compound degradation varied depending on prevailing redox conditions.
Therefore, it stands to reason that the redox changes recorded for the soil arrays over time could have triggered a change in the metabolic capabilities and composition of predominant microorganisms leading to changes in the H’ over time and depth. The general decrease in H’ with depth over the period of investigation was more pronounced in
arrays B and C; presumably the result of nutrient and electron acceptor supply and demand through the vertical profiles of the arrays. The migration of leachate pollutants through the soil profile can further contribute to lower H’ values indicative of reduced bacterial diversity. Maila, et al. (2005) came to a similar conclusion when studying the microbial diversity of different soil layers at a site polluted by hydrocarbons.
The mean H’ recorded for treatment HLRh at T1 was significantly different from that recorded at T2 and T3 (F=<0.001) but not significantly different from that recorded at T4 (Table 6.2). In contrast mean H’ recorded for treatment HLRl was significantly different (F=<0.001) at all sampling times (Table 6.2). As with mean S, the mean H’ was highest in array B, followed by arrays C, D, and A for both treatments. This is not surprising, since S and H’ are positively correlated (Hill et al., 2003).
A comparison of the mean H’ recorded for each soil array of both treatments at the comparative sampling times revealed similar trends but different levels of diversity (Figure 6.7). At each of the four sampling intervals treatment HLRl showed higher mean H’. The trend was similar to that plotted by mean S over time, i.e. an initial increase in mean H’, peaking at T2, followed by a steady decrease to T4. If H’ is expressed as eH’
(Hill et al., 2003), essentially, this implies that at T1; T2; T3; and T4 the samples reflected a mean H’ corresponding to 15; 30; 22; and 14 equally abundant bands, respectively for treatment HLRh. By comparison, the mean H’ reflected for treatment HLRl at the respective times was indicative of H’-values representing 16; 35; 28; and 19 equally abundant bands (S). Clearly, there are differences between these values and those recorded for S (Table 6.2). The difference between these calculated values and observed values was due to unevenness in the Bacterial populations comprising the samples and can be attributed to the difference in HLR observed between the two treatments.
The difference in mean H’ recorded for both treatments at T1 was not significant (P=0.629), however, significant differences between the treatments were found at T2 (P=0.002), T3 (P=0.013), and T4 (P=<0.001) using the pairwise-t-test. A consequence of these findings is that a doubling of landfill leachate supply to the soil beneath a landfill could result in a significant decrease in H’ over time. This would mean that the soil profile would become more characteristic of a pollutant perturbed system, made up of fewer, yet more pronounced, bacterial communities.
0 0.5 1 1.5 2 2.5 3 3.5 4
0 20 40 60 80 100
Sampling Time (Weeks)
Mean Shannon-Weaver Index
Figure 6.7 Changes in the Bacterial species diversity reflected by the Shannon-Weaver Index (H’) over depth and time. Symbols correspond to soil arrays A (◊), B (■), C (∆), and D (●) that were destructively sampled at times (weeks): 12 (T1); 32 (T2); 52 (T3); and 80 (T4), respectively. (a) Treatment HLRh and (b) Treatment HLRl.
Figure 6.8 Changes in the mean Shannon-Weaver Index measured at weeks 12 (T1), 32 (T2), 52 (T3), and 80 (T4) in soil arrays A, B, C, and D, respectively. Symbols correspond to treatments HLRh (▲) and HLRl (■).
1 1.5 2 2.5 3 3.5 4
0 5 10 15 20 25 30 35 40
1 1.5 2 2.5 3 3.5 4
0 10 20 30 40
Shannon-Weaver Index (H’)
Depth (cm) (a)
(b)
6.2.1.2.3 Shannon-Weaver Evenness index of diversity (EH)
The relationship between depth and EH over time for both HLR treatments are shown in Figure 6.9. The assumption with respect to EH is that the most numerically equitable community must contain equal numbers of all species comprising the community, resulting in a EH equal to one (Camargo et al., 2005). In the context of this investigation this means that the most numerically equitable sample must contain DNA bands (species), each with equal signal intensities. The changes in EH with depth revealed a stable trend for array A (T1) for both treatments. As time progresses the trend became more erratic along the soil profile depth for both treatments, indicative of changes in the microenvironment from one depth to another that contributes significantly to the changes in community evenness at each depth over time. At T1 there was a general increase in EH
with depth for both treatments, however, at the remaining sampling intervals EH reflected a general decrease with depth for both treatments with this pattern being most prominent at T4. The general decrease in EH from T1 to T4 for both treatments, suggests that there was an overall decrease in Bacterial diversity and that the conditions became favorable for the dominance of specific species.
The mean EH recorded for both treatments over all sampling times showed no overall significant differences (Table 6.2). However, one could state that the overall level of significance excluding the effect of time on community evenness as being random was much smaller for treatment HLRh (F=0.067) in comparison to treatment HLRl (F=0.856).
Evidence of this was observed in the difference between E at T1 and T4 for treatment HLRh (Table 6.2). Therefore, one could speculate that, given sufficient time, the effect of time on community evenness could become significant under the two leachate HLR treatments investigated.
A comparison of the mean EH recorded for each soil array of both treatments at the comparative sampling times revealed similar trends but different levels of community evenness (Figure 6.10). Treatment HLRh showed greater community evenness over T1 and T2. At T3 the mean EH recorded was the same for both treatments, followed by a more pronounced decrease in mean EH for treatment HLRh in comparison to HLRl. This represented a deviation from the trajectories plotted for the previous measures of community diversity (mean S and H’) where treatment HLRl reflected constantly higher
mean S and H’ over all sampling times. EH can be expressed as eH’/S (Hill et al., 2003), from this ratios of 0.86; 0.79; 0.80; and 0.76 were derived for T1, T2, T3 and T4 of the HLRh treatment. This means that of the S recorded for T1; T2; T3; and T4 (Table 6.2) of treatment HLRh, the unevenness of the mean species abundance (band density) gave each sampling time a mean value of 86 %; 79 %; 80 %; and 76 % of the expected mean EH if all the species (bands) had an equal abundance at the respective sampling times. One can arrive at a similar conclusion for treatment HLRl, where unevenness accounted for 18 %;
23 %; 20 %; and 22 % of the mean species abundance at T1; T2; T3; and T4, respectively.
There are numerous factors that contribute to the unevenness exhibited in each of the soil arrays. An investigation examining bacterial diversity by amplified ribosomal DNA restriction analysis (ARDRA) in zinc contaminated agricultural soils found decreasing bacterial diversity and associated evenness with increasing zinc pollution (Moffett, Nicholson, Uwakwe, Chambers, Harris, and Hill, 2003).
The difference in mean EH recorded for both treatments at all sampling times was not significant by virtue of the pairwise-t-test. Mean EH (Table 6.2) recorded at all sampling times and for both treatments was close to unity, indicating an even yet numerically dynamic distribution of members of the Bacterial community over time.
0.9 0.92 0.94 0.96 0.98
0 20 40 60 80 100
Sampling Time (Weeks) Mean Shannon-Weaver Eveness Index
Figure 6.9 Changes in the Bacterial species evenness reflected by the Shannon-Weaver Evenness Index (EH’) over depth and time. Symbols correspond to soil arrays A (◊), B (■), C (∆), and D (●) destructively sampled at times (weeks): 12 (T1); 32 (T2); 52 (T3); and 80 (T4), respectively. (a) Treatment HLRh and (b) Treatment HLRl.
Figure 6.10 Changes in the mean Shannon-Weaver Index measured at weeks 12 (T1), 32 (T2), 52 (T3), and 80 (T4) in soil arrays A, B, C, and D, respectively. Symbols correspond to treatments HLRh (▲) and HLRl (■).
0.8 0.85 0.9 0.95 1 1.05
0 5 10 15 20 25 30 35 40
0.8 0.85 0.9 0.95 1
0 5 10 15 20 25 30 35 40
Shannon-Weaver Evenness Index (EH)
Depth (cm) (b)
(a)
Table 6.3 The significance of P-values of the two sample pair-wise t-test for comparing the Species Richness (S), Shannon-Weaver Index (H’), and the Shannon-Weaver Evenness Index (E) for two different hydraulic loading rates (HLR) over time
P-value Factors
Compared Time 1 Time 2 Time 3 Time 4
SH vs SL 0.283 0.002 0.002 <0.001
H’H vs H’L 0.629 0.002 0.013 <0.001
EH vs EL 0.055 0.356 0.926 0.116
6.2.1.3 The effects of the redox, pH, and phenol concentration of landfill leachate on